In the description which follows, references are made to certain literature citations which are listed at the end of the specification and all of which are incorporated herein by reference.
Transmembrane proteins have been classified in several major classes, including G protein coupled receptors, transporters, tyrosine kinase receptors, cytokine receptors and LDL receptors. G protein coupled receptors (GPCRs) can be grouped on the basis of structure and sequence homology into several families. Family 1 (also referred to as family A or the rhodopsin-like family) is by far the largest subgroup and contains receptors for small molecules such as the catecholamines, dopamine and noradrenaline, peptides such as the opioids, somatostatin and vasopressin, glycoprotein hormones such as thyrotropin stimulating hormone and the entire class of odorant molecules (George et al, 2002). Family 2 or family B contains the receptors such as for glucagon, parathyroid hormone and secretin. These GPCRs are characterised by a long amino terminus that contains several cysteines, which may form disulphide bridges. Family 3 or family C contains receptors such as the metabotropic glutamate, the Ca2+-sensing and the gamma-amino butyric acid (GABA)B receptors. These receptors are also characterised by a complex amino terminus. Although all GPCRs share the seven membrane-spanning helices, the various GPCR families show no sequence homology to one another.
GPCRs are the largest known group of cell-surface mediators of signal transduction and are present in every cell in the body. GPCR action regulates the entire spectrum of physiological functions, such as those involving the brain, heart, kidney, lung, immune and endocrine systems. Extensive efforts during the past decade has identified a large number of novel GPCRs, including multiple receptor subtypes for previously known ligands, and numerous receptors for which the endogenous ligands are as yet unidentified, termed ‘orphan’ receptors or oGPCRs (Lee et al., 2001; Lee et al., 2002; Bailey et al., 2001).
GPCRs have been the successful targets of numerous drugs for diverse disorders in clinical use today, with an estimated 50% of the current drug market targeting these molecules. Among the known GPCRs, ˜335 receptors are potential drug development targets, of which 195 have known ligands, and the remaining 140 being oGPCRs, awaiting identification of their ligands. Although various methodological advances have accelerated the pace of novel receptor discovery, the pace of ligand and drug discovery lags far behind. Conventional, small-scale pharmacological screening assay methods were initially used to discover the ligands and drugs for many of the GPCRs, but newer assay procedures are continually being sought.
Since GPCRs form over 80% of cell surface receptors, they represent a substantial resource and constitute a highly relevant group of protein targets for novel drug discovery. Drugs interacting with GPCRs have the potential to be highly selective, as the interactions will be confined to the cell surface and to tissues bearing the receptors exclusively. The convergence of the discovery of GPCRs with the realisation that they are important drug targets, has led to intense pharmaceutical interest in devising better ways to detect and screen for compounds interacting with GPCRs. Therefore, creating improved assay methods is an urgent requirement towards the goal of more rapid drug screening and discovery. There is a need to optimise the ability to detect an interaction between test compounds and the receptors, which is the fundamental initial step in the process of drug development.
Improved ligand-identification strategies to accelerate the characterisation of all GPCRs will define their physiological functions and realise their potential in discovering novel drugs. Even with the identified GPCRs, there is a paucity of highly selective subtype specific drugs being discovered and pharmaceutical houses are experiencing a dearth of promising lead compounds, in spite of the wealth of drug targets defined. The list of new drug product approvals by the top 20 pharmaceutical companies has declined considerably over the period 1999-2001, compared to the preceding three year period (Smith, 2002). Thus there is a real need to have improved, versatile assay systems, where not just endogenous ligands, but novel compounds interacting with receptors can be tested and identified in a quick and efficient manner that is amenable to automation.
As the signal transduction pathway required to activate an oGPCR cannot be predicted, an assay system for interacting compounds which is independent of prior predictions of which effector system (such as adenylyl cyclase, PLC, cGMP, phosphodiesterase activity) is employed by the receptor is required. Assigning ligands to GPCRs and oGPCRs is an important task; however the diversity of both GPCR ligands and effector systems can limit the utility of some existing ligand-identification assays, requiring novel approaches to drug discovery.
Recently, several methods utilising refined assay systems testing tissue extracts, large ligand libraries and specific ligands of interest have successfully discovered the endogenous ligands for a number of these oGPCRs. Such methods have been collectively referred to as “Reverse Pharmacology” (Howard et al., 2001). Various methods have been used to assay induced cell activity in response to an agonist compound, including the Fluorescence Imaging Plate Reader assay (FLIPR, Molecular Devices Corp., Sunnyvale, Calif.) and Barak et al., (1997), and U.S. Pat. Nos. 5,891,646 and 6,110,693 which disclose the use of a β-arrestin-green fluorescent fusion protein for imaging arrestin translocation to the cell surface upon stimulation of a GPCR.
The potential disadvantages of such methods are as follows: 1) visualisation is not of the receptor; 2) the protein translocation requires complex computerised analytical technologies; 3) prior identification of agonist is necessary to screen for antagonists; and 4) specific G protein coupling is necessary to generate a signal.
Mechanisms of ligand binding and signal transduction by GPCRs traditionally have been modelled on the assumption that monomeric receptors participate in the process, and a monomeric model for GPCRs has been generally accepted. Since the mid-1990s, however, numerous reports have demonstrated oligomerisation of many GPCRs (reviewed by George et al., 2002), and it is now realised that oligomerisation is an inherent aspect of GPCR structure and biology. Also certain receptor subtypes formed hetero-oligomers, and these receptors have functional characteristics that differ from homogeneous receptor populations. At present, studies of GPCR oligomerisation do not make a distinction between dimers and larger complexes, and the term dimer is used interchangeably with the terms oligomer and multimer. There are no conclusive data to indicate how large the oligomers of functional GPCRs are. Importantly, generation of new properties through hetero-oligomerisation suggested a mechanism for generating diversity of function among GPCRs. Homooligomerisation of GPCRs is accepted as a universal occurrence and a number of GPCRs are known to assemble as heterooligomeric receptor complexes (George et al., 2002). For example, the GABA-B1 and GABA-B2 receptors are not functional individually and only form a functional receptor when co-expressed (White et al., 1998). The assembly of heterooligomer receptor complexes can result in novel receptor-ligand binding, signalling or intracellular trafficking properties. For example, co-transfection of the mu and delta opioid receptors resulted in the formation of oligomers with functional properties that were distinct from each of the receptors individually (George et al., (2000)). The interaction of mu and delta opioid receptors to form oligomers generated novel pharmacological and G protein coupling properties. When mu and delta opioid receptors were co-expressed, the highly selective agonists (DAMGO, DPDPE, and morphine) had reduced potency and altered rank order, whereas certain endogenous ligands endomorphin-1 and Leu-enkephalin had enhanced affinity, suggesting the formation of a novel ligand binding pocket (George et al., 2000). In contrast to the individually expressed mu and delta receptors, the coexpressed receptors showed pertussis toxin insensitive signal transduction, likely due to interaction with a different subtype of G protein. It would therefore be very useful, from the point of view of identification of potential drug targets, to have a means of determining whether a particular pair of GPCRs are able to form heterooligomers.
In many reports, heterooligomers have been tentatively identified by the ability to co-immunoprecipitate. When two GPCRs are shown to co-immunoprecipitate, however, there are two possible interpretations; either the receptors are directly physically interacting, or both are interacting through contact with a common third protein (or proteins). An alternative approach to detecting receptor oligomers has been the development of energy transfer assays using bioluminescent resonance energy transfer (BRET) or fluorescence resonance energy transfer (FRET). Although these methods detect energy transfer between two receptor molecules labelled by fluorophores at proximities of less than 100 angstroms, it is unclear whether receptor conformational changes can be reliably distinguished from de novo oligomerisation.
Transporters are protein pumps that move molecules, ions and other chemicals in and out of cells and exist in virtually all cells. The transporters can be grouped into families on the basis of structure, sequence homology and the molecules they transport. Separate transporters exist for monoamine neurotransmitters such as dopamine, serotonin, norepinephrine and GABA, for amino acids such as glycine, taurine, proline and glutamate, for vesicular monoamines, acetylcholine and GABA/glycine, for sugars such as glucose and disaccharides, for organic cations and organic anions, for oligopeptides and peptides, for fatty acids, bile acids, nucleosides, for water and for creatine. Pumps that export large molecules such as drugs, toxins and antibiotics from the cell are exemplified by the P-glycoprotein (multidrug resistance protein) family. There are also several related transporters the function of which remains unknown (Masson et al., 1999). These transporters are membrane proteins consisting of a polypeptide generally with 12 transmembrane domains. The glutamate and aspartate transporters belong to a separate family whose members have 6 to 10 TM domains and share no homology to the other transporters (Masson et al., 1999). Both the amino and carboxyl termini are located on the intracellular side of the membrane.
A large number of neurological and psychiatric disorders including depression, Parkinson's disease, schizophrenia, drug addiction, Tourette's syndrome, and attention deficit disorders are considered to involve the monoamine transporters. The dopamine transporter (DAT) is the major target for psychostimulants such as cocaine and methylphenidate. The transporters have been the successful targets of numerous drugs for diverse disorders in clinical use today, particularly antidepressant drugs, including fluoxetine, sertraline and the other related serotonin selective reuptake inhibitors (SSRIs). Although methodological molecular advances have identified the known transporters, the pace of ligand and drug discovery lags behind. Conventional, pharmacological screening assay methods were used to discover the ligands and drugs for some of the transporters, but newer assay procedures are urgently being sought. Improved ligand-identification strategies to accelerate the characterisation of all the transporters will further define their physiological functions and realise their potential in discovering novel drugs. Even with the identified transporters, there is a paucity of highly selective specific drugs being discovered.
The tyrosine kinase receptor family members are characterised by their structural similarity, with an extracellular ligand binding domain, a single transmembrane domain and an intracellular domain with tyrosine kinase activity for signal transduction. There are many subfamilies of receptor tyrosine kinases, exemplified by the epidermal growth factor (EGF) receptor (also called HER1 or erbB1), which is one of four members of such a subfamily, which also includes HER2, HER3 and HER4. The principal EGF-R ligands are EGF, TGF-α, heparin binding EGF, amphiregulin, betacellulin and epiregulin (Shawver et al., 2002). Activation of the EGF-R causes the receptor to dimerise with either another EGF-R monomer or another member of the HER subfamily. Marked diversity of ligand binding and signalling is generated by the formation of heterodimers among family members (Yarden and Sliwkowski, 2001). The EGF-R is widely expressed in a variety of tissues and mediates important functions such as cell growth and tissue repair. Overexpression of EGF-R occurs in many types of cancer, such as head and neck, lung, laryngeal, esophageal, gastric, pancreatic, colon, renal, bladder, breast, ovarian, cervical, prostate, thyroid, melanoma and glioma, and correlates with a poor outcome (Nicholson et al., 2001). Therefore there is great interest and need for developing drugs targeting the EGF-R and for methods which assist in identifying such drugs.
Other subfamilies of receptor tyrosine kinases are exemplified by the receptors for vascular endothelial factor (four members) and fibroblast growth factor (four members). These have important roles in angiogenesis and also have significant roles in the uncontrolled proliferation of vessels characterizing carcinogenesis (Hanahan and Folkman. 1996).
The cytokine receptors are proteins spanning the membrane with an extracellular ligand binding domain and an intracellular domain with intrinsic kinase activity or adapter regions able to interact with intracellular kinases. The receptors are divided into subclasses based on their structural complexity. The ‘simple’ receptors are those including receptors for growth hormone, erythropoietin and interleukins, and the ‘complex’ receptors include the tumour necrosis factor receptor family, the 4-helical cytokine receptor family, the insulin/insulin-like receptor family and granulocyte colony stimulating receptor (Grotzinger, 2002).
The insulin and insulin-like growth factor (IGF) receptor family controls metabolism, reproduction and growth (Nakae et al., 2001). There are nine different insulin-like peptides known and there are three known receptors that interact with them, IR, IGF-1R and IGF-2R, and an orphan member IR-related receptor. Each receptor exists as homodimers on the cell surface or heterodimers. The IR subfamily is also related to the EGF-R family.
IR, produced from a single mRNA, undergoes cleavage and dimerisation and translocation to the plasma membrane. Each monomer component contains a single transmembrane domain; the complete receptor comprised two α and two β subunits, linked by disulphide bridges. The β subunit contains the single TM and the intracellular region. This receptor is a tyrosine kinase that catalyzes the phosphorylation of several intracellular substrates.
The low density lipoprotein (LDL)-receptor family act as cargo transporters, regulating the levels of lipoproteins and proteases (Strickland et al., 2002). There are nine recognised members of the family, all of which share structural similarity, including an extracellular region, a single transmembrane domain region and a cytoplasmic tail. The LDL receptor plays a major role in the clearance of lipoproteins, and genetic defects in the LDL receptor can result in the accumulation of LDL in the bloodstream.
The first characterised motif shown to be able to direct protein nuclear importation was exemplified by the amino acid sequence (PKKKRKV: SEQ ID NO: 129) contained in the SV40 large T antigen protein. The nuclear localisation sequence (NLS) motifs are recognised by the importin α-βreceptor complex, which binds the NLS (Gorlich et al., 1996). These are cytosolic proteins, which recognise NLS containing proteins and transport these proteins to dock at the nuclear pore. The entire complex subsequently docks at the nuclear pore complex (Weis et al., 1998, Schlenstedt et al., 1996), contained at the nuclear envelope. The nuclear envelope is a boundary containing pores that mediate the nuclear transport process (Weis et al., 1998).
There have been very few and rare reports of GPCRs localising in the nucleus. One such example is the GPCR angiotensin type 1 (AT1) receptor, which contains an endogenous NLS which serves to direct the GPCR into the nucleus (Lu et al., 1998), providing evidence that this NLS sequence was involved in the nuclear targeting of the AT1 receptor. These authors and Chen et al., (2000) reported that AT1 receptors increased in the nucleus in response to agonist. The nuclear localisation of the parathyroid hormone receptor has been reported (Watson et al, 2000). However very few of the superfamily of GPCRs contain an endogenous NLS mediating translocation of the receptor to the nucleus.
There therefore remains a need for new, more convenient methods for identifying compounds which interact with transmembrane proteins such as GPCRs, transporters, etc. There also remains a need for improved, less ambiguous methods for detecting oligomerisation of transmembrane proteins.